You Can Have My Class AB When You Pry It From My Cold Dead Fingers

[DaveVoorhis]

Push pull inherently implies AC operation.

Yes, but the converters are DC, notionally equivalent to the +DC and -DC power rails in a conventional Class A or AB amplifier. In a conventional Class A or AB amplifier, the fixed-voltage DC power rails are modulated by the output devices to deliver AC to the speakers. In this topology, the DC power rails (for the speaker only) are provided by DC-to-DC converters whose output voltage is modulated by the output devices (in this case, tubes) to deliver AC to the speakers.

 

[Pio1980]

Push pull inherently implies AC operation.

Push-pull, aka differential.

 

[JoeESP9]

A differential amplifier amplifies the difference between two AC signals. The result is an AC output no matter how you want to obfuscate things. Push pull and differential are two completely different things.

 

[Pio1980]

Yes, but the converters are DC, notionally equivalent to the +DC and -DC power rails in a conventional Class A or AB amplifier. In a conventional Class A or AB amplifier, the fixed-voltage DC power rails are modulated by the output devices to deliver AC to the speakers. In this topology, the DC power rails (for the speaker only) are provided by DC-to-DC converters whose output voltage is modulated by the output devices (in this case, tubes) to deliver AC to the speakers.

A very interesting approach, deserving of a plain language description. Thanx for anything you can provide. The copy also mentions screen drive.

 

[Pio1980]

A differential amplifier amplifies the difference between two AC signals. The result is an AC output no matter how you want to obfuscate things

I'm not obfuscating anything, I'm trying to understand the tech. And differential operation applies to DC as well as AC.

 

[Powertech]

Pulsating (modulated) DC will not drive any speaker I'm aware of. There has to be a conversion to AC. To drive any unpowered speaker I'm aware of you need AC.

Nope.
Let's try to explain this. Please bear with me - this is going somewhere.
In a variable speed (AKA variable frequency) AC motor drive starts off with an AC to DC rectifier. The DC bus feeds a set of triacs (solid state DC switches). The motor does not receive a sine wave as it would from the mains supply. What it actually receives is a high speed (usually around 4kHz) series of DC pulses which represent a sine wave. The pulses are created by a driver circuit which triggers high speed triac switching devices. The rate, width and amplitude of each pulse is variable. The result is received by the motor coils and approximated such that the motor behaves as though it is receiving a variable frequency sine wave. I know how this works as I used to commission and service these systems.

With an audio signal the same process can be used, albeit at a much higher switching frequency - in the megahertz range in this case. The required output waveform can be produced by varying the amplitude, pulse width and pulse rate such that the resultant signal energy replicates the audio signal used as the control signal for the pulse generator circuit. The loudspeaker cannot react to the pulses in the megahertz range so it averages out the amplitude of the high speed pulses that it does see. With a bit of digital wizardry this averaging can be made to appear like a music signal to the loudspeaker coil which then translates this into the same motion as it would with a true music signal. Think about how A to D and D to A converters work and then make it much more powerful and you would have a similar effect.

 

[DaveVoorhis]

A differential amplifier amplifies the difference between two AC signals. The result is an AC output no matter how you want to obfuscate things

There's no intention to obfuscate. Quite the opposite!

Per the patent, there are two DC-to-DC converters, one supplying a +DC voltage and the other supplying a -DC voltage. The output of each converter is controlled by a tube, acting as a relatively conventional tube amplifier. The input signal is applied directly to the + side converter's tube grid, and is inverted and applied to the - side converter tube's grid.

The result is that a varying AC input signal voltage produces a varying AC output voltage to the speakers, by varying the level of +DC and -DC emitted by the pair of converters.

 

[DaveVoorhis]

Nope.
Let's try to explain this. Please bear with me - this is going somewhere.
In a variable speed (AKA variable frequency) AC motor drive starts off with an AC to DC rectifier. The DC bus feeds a set of triacs (solid state DC switches). The motor does not receive a sine wave as it would from the mains supply. What it actually receives is a high speed (usually around 4kHz) series of DC pulses which represent a sine wave. The pulses are created by a driver circuit which triggers high speed triac switching devices. The rate, width and amplitude of each pulse is variable. The result is received by the motor coils and approximated such that the motor behaves as though it is receiving a variable frequency sine wave. I know how this works as I used to commission and service these systems.

With an audio signal the same process can be used, albeit at a much higher switching frequency - in the megahertz range in this case. The required output waveform can be produced by varying the amplitude, pulse width and pulse rate such that the resultant signal energy replicates the audio signal used as the control signal for the pulse generator circuit. The loudspeaker cannot react to the pulses in the megahertz range so it averages out the amplitude of the high speed pulses that it does see. With a bit of digital wizardry this averaging can be made to appear like a music signal to the loudspeaker coil which then translates this into the same motion as it would with a true music signal. Think about how A to D and D to A converters work and then make it much more powerful and you would have a similar effect.

That's the basis for Class D amplification. The most recent discussion is about the David Berning "Quadrature Z", which is notionally Class A or AB depending on how the output tube is biased, but drives a DC-to-DC converter instead of driving the speaker directly, thus providing impedance matching without the usual heavy-iron audio output transformer. See http://davidberning.com/products/qz and http://davidberning.com/technology/patent5612646

 

[JoeESP9]

There's no intention to obfuscate. Quite the opposite!

Per the patent, there are two DC-to-DC converters, one supplying a +DC voltage and the other supplying a -DC voltage. The output of each converter is controlled by a tube, acting as a relatively conventional tube amplifier. The input signal is applied directly to the + side converter's tube grid, and is inverted and applied to the - side converter tube's grid.

The result is that a varying AC input signal voltage produces a varying AC output voltage to the speakers, by varying the level of +DC and -DC emitted by the pair of converters.

I believe I already said that AC in results in AC out.

That's the basis for Class D amplification. The most recent discussion is about the David Berning "Quadrature Z", which is notionally Class A or AB depending on how the output tube is biased, but drives a DC-to-DC converter instead of driving the speaker directly, thus providing impedance matching without the usual heavy-iron audio output transformer. See http://davidberning.com/products/qz and http://davidberning.com/technology/patent5612646

I also said the Berning was not Class D.

 

[DaveVoorhis]

I believe I already said that AC in results in AC out.

That's true of every audio amplifier. I was attempting to clarify what appeared to be your misunderstanding in post #74. Sorry if I misunderstood your question, but hopefully the explanations will help others.

 

[Pio1980]

The DC-DC converter is switchmode, is it not the effective power stage?

 

[DaveVoorhis]

The DC-DC converter is switchmode, is it not the effective power stage?

I'd describe it as a variable voltage power supply.

 

[Pio1980]

I'd describe it as a variable voltage power supply.

Whatever it seems, it is ingenious.
I haven't encountered any of his products in the wild as of yet, one of the amps should be on my bucket list.

 

[JoeESP9]

The DC-DC converter is switchmode, is it not the effective power stage?

Since there are two DC to DC converters, one +DC and one -DC when they are coupled together the result is AC.

 

[DaveVoorhis]

Whatever it seems, it is ingenious.

Yes, it's very clever and nicely gets around the usual high tube count needed for traditional OTL tube circuits. I'd like to hear it.

 

[JoeESP9]

Yes, it's very clever and nicely gets around the usual high tube count needed for traditional OTL tube circuits. I'd like to hear it.

So would I. However, it doesn't have enough power for my purposes.

 

[45rpmspinner]

While I'm fine with accepting the opinion of those who know better, my little brain has trouble getting past the fact that the Berning amp utilizes a 250khtz carrier wave.
In general: "mind blown".

 

[DaveVoorhis]

While I'm fine with accepting the opinion of those who know better, my little brain has trouble getting past the fact that the Berning amp utilizes a 250khtz carrier wave.

Same principle as AM radio: audio modulation of an RF carrier. Though in this case, the 250khz "carrier" appears to only be used to support DC-to-DC variable voltage conversion.

 

[nj pheonix]

 

[Archguy]

Interesting that Luxman cagily shows no power consumption in the back apron cheesecake photo

Thanks for your help. I've lately noticed a number of mfrs/retailers who display photo-edited images.